Transformer fault prediction method and device, terminal and storage medium

Abstract

The invention provides a transformer fault prediction method and device, a terminal and a storage medium. The method comprises the following steps: acquiring various target monitoring data at different acquisition moments at different positions of a target transformer; taking each type of target monitoring data as a characteristic dimension of the target transformer, and constructing a multi-dimensional tensor for the multiple types of target monitoring data according to a time dimension, a space dimension and the characteristic dimension; segmenting the multi-dimensional tensor according to a time dimension, a space dimension and a feature dimension, and extracting segmented space dimension and feature dimension slice data, time dimension and feature dimension slice data and relevance of the time dimension and space dimension slice data; and obtaining a fault prediction result of the target transformer according to an extraction result. According to the method, the relevance of various kinds of target monitoring information of the target transformer among all dimensions can be fully extracted, and then the accuracy of transformer fault prediction is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of fault prediction, and in particular to a transformer fault prediction method, device, terminal and storage medium. Background Art

[0002] As a key equipment in the power system, the operating status of power transformers directly affects the safety and stability of the power grid. However, transformer failures often involve complex factors such as electromagnetic, mechanical, and chemical factors. Traditional single monitoring methods are difficult to fully reflect their true operating status. Existing fault prediction methods, such as the IEC three-ratio method and infrared detection method, have their own advantages, but they all have limitations, such as insufficient accuracy and poor universality. Therefore, it is particularly important to develop a new method that can integrate multiple monitoring data and improve the accuracy of fault prediction. Summary of the invention

[0003] The embodiments of the present invention provide a transformer fault prediction method, device, terminal and storage medium to solve the problem that the current transformer fault prediction methods have limitations and the fault prediction accuracy is not high.

[0004] In a first aspect, an embodiment of the present invention provides a transformer fault prediction method, comprising:

[0005] Acquire multiple target monitoring data at different locations of the target transformer at different collection times;

[0006] Taking each of the target monitoring data as a feature dimension of the target transformer, constructing a multidimensional tensor of the multiple target monitoring data according to the time dimension, space dimension and feature dimension;

[0007] The multidimensional tensor is divided according to the time dimension, the space dimension and the feature dimension respectively to obtain the space dimension and feature dimension slice data, the time dimension and feature dimension slice data and the time dimension and space dimension slice data of the target transformer;

[0008] The spatial dimension and feature dimension slice data, the time dimension and feature dimension slice data, and the correlation between the time dimension and spatial dimension slice data are extracted respectively, and the fault prediction result of the target transformer is obtained according to the extraction result.

[0009] In a possible implementation, the spatial dimension and feature dimension slice data, the time dimension and feature dimension slice data, and the correlation between the time dimension and spatial dimension slice data are extracted respectively, and the fault prediction result of the target transformer is obtained according to the extraction result, including:

[0010] Extracting the correlation between the spatial dimension and the feature dimension slice data, recorded as the first correlation;

[0011] Extracting the correlation between the slice data of the time dimension and the feature dimension, and recording it as the second correlation;

[0012] Extracting the correlation between the slice data of the time dimension and the space dimension, recorded as the third correlation;

[0013] The first correlation, the second correlation and the third correlation are input into a fully connected network to obtain a fault prediction result of the target transformer.

[0014] In a possible implementation, extracting the correlation between the spatial dimension and the feature dimension slice data, recorded as the first correlation, includes:

[0015] The correlation between the spatial dimension and feature dimension slice data is extracted based on the graph convolutional network and recorded as the first correlation.

[0016] In a possible implementation, extracting the correlation between the time dimension and the feature dimension slice data, recorded as the second correlation, includes:

[0017] The correlation between the time dimension and feature dimension slice data is extracted based on the gated recurrent unit and recorded as the second correlation.

[0018] In a possible implementation, extracting the correlation between the slice data of the time dimension and the space dimension, recorded as the third correlation, includes:

[0019] The correlation between the slice data of the time dimension and the space dimension is extracted based on the spatiotemporal graph convolutional network and recorded as the third correlation.

[0020] In a possible implementation, after acquiring a plurality of target monitoring data at different locations of the target transformer at different acquisition times, the method further includes:

[0021] Performing data cleaning and data standardization on each of the target monitoring data to obtain pre-processed target monitoring data;

[0022] The target monitoring data are constructed into a multidimensional tensor according to the time dimension, space dimension and feature dimension, including:

[0023] A variety of pre-processed target monitoring data are constructed into multi-dimensional tensors according to the time dimension, space dimension and feature dimension.

[0024] In a second aspect, an embodiment of the present invention provides a transformer fault prediction device, comprising:

[0025] An acquisition module, used to acquire a variety of target monitoring data at different locations of a target transformer at different acquisition times;

[0026] A first processing module is used to use each of the target monitoring data as a feature dimension of the target transformer, and construct a multidimensional tensor of the multiple target monitoring data according to the time dimension, space dimension and feature dimension;

[0027] The second processing module is used to slice the multidimensional tensor according to the time dimension, the space dimension and the feature dimension respectively, to obtain the space dimension and feature dimension slice data, the time dimension and feature dimension slice data and the time dimension and space dimension slice data of the target transformer;

[0028] A prediction module is used to respectively extract the spatial dimension and feature dimension slice data, the time dimension and feature dimension slice data, and the correlation between the time dimension and spatial dimension slice data, and obtain a fault prediction result of the target transformer based on the extraction result.

[0029] In a possible implementation, the prediction module is specifically used to:

[0030] Extracting the correlation between the spatial dimension and the feature dimension slice data, recorded as the first correlation;

[0031] Extracting the correlation between the slice data of the time dimension and the feature dimension, and recording it as the second correlation;

[0032] Extracting the correlation between the slice data of the time dimension and the space dimension, recorded as the third correlation;

[0033] The first correlation, the second correlation and the third correlation are input into a fully connected network to obtain a fault prediction result of the target transformer.

[0034] In a third aspect, an embodiment of the present invention provides a terminal, comprising a memory and a processor, wherein the memory is used to store a computer program, and the processor is used to call and run the computer program stored in the memory to perform the steps of the method described in the first aspect or any possible implementation of the first aspect.

[0035] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method described in the first aspect or any possible implementation method of the first aspect are implemented.

[0036] The embodiment of the present invention provides a transformer fault prediction method, device, terminal and storage medium, which first obtains multiple target monitoring data at different acquisition times at different positions of the target transformer, then uses each target monitoring data as a characteristic dimension of the target transformer, and constructs a multidimensional tensor according to the time dimension, space dimension and feature dimension of the multiple target monitoring data; then the multidimensional tensor is divided according to the time dimension, space dimension and feature dimension respectively, and the space dimension and feature dimension slice data, the time dimension and feature dimension slice data and the time dimension and space dimension slice data of the target transformer are obtained; then the correlation between the space dimension and feature dimension slice data, the time dimension and feature dimension slice data and the time dimension and space dimension slice data are extracted respectively, and the fault prediction result of the target transformer is obtained according to the extraction result. Therefore, when obtaining the fault prediction result of the target transformer, the multiple target monitoring data of the target transformer can be integrated, and the correlation between the multiple target monitoring information of the target transformer in each dimension can be fully extracted, thereby improving the accuracy of transformer fault prediction. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0038] Figure 1 is a flowchart of the transformer fault prediction method provided by an embodiment of the present invention;

[0039] Figure 2 It is a technical roadmap of the transformer fault prediction method provided by an embodiment of the present invention;

[0040] Figure 3 It is a schematic diagram of dividing a multidimensional tensor according to a time dimension, a space dimension, and a feature dimension, respectively, provided by an embodiment of the present invention;

[0041] Figure 4 is a schematic diagram of the structure of a transformer fault prediction device provided by an embodiment of the present invention;

[0042] Figure 5 is a schematic diagram of a terminal provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0043] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present invention. However, it should be clear to those skilled in the art that the present invention may be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present invention.

[0044] In order to make the purpose, technical solutions and advantages of the present invention more clear, specific embodiments will be described below in conjunction with the accompanying drawings.

[0045] See also Figure 1 and Figure 2 , which shows a flowchart of the transformer fault prediction method provided by an embodiment of the present invention, and is described in detail as follows:

[0046] In step 101, a variety of target monitoring data at different locations of a target transformer at different collection times are acquired.

[0047] In this embodiment, in order to improve the accuracy of transformer fault prediction, multiple monitoring data of the integrated transformer are considered. Therefore, when it is necessary to predict the fault of the target transformer, multiple monitoring data at different locations of the target transformer at different collection times are obtained and recorded as multiple target monitoring data.

[0048] For example, the various monitoring data of the transformer obtained may include oil temperature, dissolved gas content in the oil (such as hydrogen H 2 、Methane CH 4 , ethane C 2 H 6 A variety of sensors can be set at different locations of the transformer to collect multi-dimensional status monitoring data of the transformer over a period of time in history.

[0049] For example, assuming that five monitoring points are set on the transformer, an oil temperature sensor, a gas content detection sensor, a current sensor, a voltage sensor, etc. can be set at each monitoring point of the transformer, so as to facilitate subsequent fault prediction by integrating multiple monitoring data at each collection moment at each position of the transformer.

[0050] In one embodiment, after acquiring a plurality of target monitoring data at different locations of the target transformer at different acquisition times, the following may also be included:

[0051] Data cleaning and data standardization are performed on each type of target monitoring data to obtain preprocessed target monitoring data.

[0052] In this embodiment, after obtaining multiple monitoring data (that is, multiple target monitoring data of the target transformer), data cleaning processing and data standardization processing are performed to eliminate noise and abnormal values ​​in various target monitoring data, which is conducive to more accurate fault prediction in the subsequent process.

[0053] In step 102, each target monitoring data is used as a feature dimension of the target transformer, and multiple target monitoring data are constructed into a multidimensional tensor according to the time dimension, space dimension and feature dimension.

[0054] In one embodiment, a plurality of target monitoring data are constructed into a multidimensional tensor according to time dimension, space dimension and feature dimension, and a plurality of preprocessed target monitoring data are constructed into a multidimensional tensor according to time dimension, space dimension and feature dimension.

[0055] In this embodiment, considering that when currently predicting transformer faults, even if multiple monitoring data are considered, the correlation between different monitoring data is often ignored, or the correlation between different monitoring data cannot be fully extracted, resulting in the transformer fault prediction result being difficult to meet expectations.

[0056] Therefore, after acquiring a variety of target monitoring data at different locations of the target transformer at different collection times, this embodiment constructs a dynamic tensor fusion network to facilitate the subsequent full extraction of the correlation between different transformer monitoring data in different dimensions.

[0057] Among them, when constructing the dynamic tensor fusion network, considering that for the corresponding transformer, different types of monitoring data such as oil temperature, dissolved gas content in oil, current, voltage, etc. correspond to a characteristic dimension of the transformer respectively, and the oil temperature, dissolved gas content in oil, current, voltage, etc. at different positions correspond to a spatial dimension of the transformer respectively, and the oil temperature, dissolved gas content in oil, current, voltage, etc. at different collection times correspond to a time dimension of the transformer respectively. Therefore, multiple target monitoring data can be constructed into a multidimensional tensor according to the time dimension, space dimension and characteristic dimension.

[0058] For example, the state of the target transformer is characterized by four monitoring data, namely, oil temperature, dissolved gas content in oil, current, and voltage. Five monitoring points are set on the target transformer, and sensor data is collected every 15 seconds. The target transformer has 4 feature dimensions, 5 spatial dimensions, and a time dimension every 15 seconds. Therefore, the sensor data collected at different locations and different collection times of the target transformer can be numbered according to the feature dimension, spatial dimension, and time dimension to form a multi-dimensional tensor of the target transformer.

[0059] In step 103, the multidimensional tensor is segmented according to the time dimension, space dimension and feature dimension respectively to obtain the space dimension and feature dimension slice data, the time dimension and feature dimension slice data and the time dimension and space dimension slice data of the target transformer.

[0060] In this embodiment, in order to fully extract the correlation between different dimensions of the transformer, after constructing the multidimensional tensor, it can be sliced ​​along the time dimension, space dimension and feature dimension respectively.

[0061] like Figure 2 and Figure 3 As shown in (a), (b), and (c), χ represents a multidimensional tensor, T represents a time dimension, N represents a space dimension, and C represents a feature dimension. When the multidimensional tensor χ is sliced ​​along the time dimension T, the spatial dimension and feature dimension slice data N×C are obtained. When the multidimensional tensor χ is sliced ​​along the space dimension N, the time dimension and feature dimension slice data T×C are obtained. When the multidimensional tensor χ is sliced ​​along the feature dimension C, the time dimension and space dimension slice data N×T are obtained.

[0062] In step 104, the spatial dimension and feature dimension slice data, the time dimension and feature dimension slice data, and the correlation between the time dimension and spatial dimension slice data are extracted respectively, and the fault prediction result of the target transformer is obtained according to the extraction result.

[0063] Exemplarily, respectively extracting the correlation between the spatial dimension and feature dimension slice data, the time dimension and feature dimension slice data, and the time dimension and spatial dimension slice data, and obtaining the fault prediction result of the target transformer according to the extraction result may include:

[0064] The correlation between the slice data of the spatial dimension and the feature dimension is extracted and recorded as the first correlation.

[0065] The correlation between the slice data of the time dimension and the feature dimension is extracted and recorded as the second correlation.

[0066] The correlation between the slice data of the time dimension and the space dimension is extracted and recorded as the third correlation.

[0067] The first correlation, the second correlation and the third correlation are input into the fully connected network to obtain the fault prediction result of the target transformer.

[0068] Exemplarily, the correlation between spatial dimension and feature dimension slice data can be extracted based on the graph convolutional network, which is recorded as the first correlation; the correlation between time dimension and feature dimension slice data can be extracted based on the gated recurrent unit, which is recorded as the second correlation; the correlation between time dimension and spatial dimension slice data can be extracted based on the spatiotemporal graph convolutional network, which is recorded as the third correlation.

[0069] In this embodiment, after obtaining the spatial dimension and feature dimension slice data, the time dimension and feature dimension slice data, and the time dimension and spatial dimension slice data, considering that there is correlation between each dimension and the correlation features are different, different networks can be used to extract the correlation between each dimension according to the characteristics of each dimension. For example, the correlation between the spatial dimension-feature dimension slice data is extracted based on the graph convolution network, the correlation between the time dimension-feature dimension slice data is extracted based on the gated recurrent unit, and the correlation between the time dimension-space dimension slice data is extracted based on the spatiotemporal graph convolution network, and then the correlation features of the three slices are input into the fully connected network for fault prediction. In this way, fault prediction is performed after fully extracting the correlation between the various dimensional information of the transformer, thereby improving the accuracy of the transformer fault prediction.

[0070] Exemplarily, before the correlation features of the three slices are input into the fully connected network for fault prediction, multiple monitoring data of the faulty transformer can be obtained, a multi-dimensional tensor can be constructed, sliced, correlation extracted and input into the fully connected network according to steps 101 to 104, thereby determining a fault threshold. Then, the correlation features of the three slices are input into the fully connected network, and the correlations of the multiple target monitoring data of the target transformer in different dimensions can be compared with the correlations of the multiple monitoring data of the faulty transformer in different dimensions, thereby obtaining the fault prediction result of the target transformer.

[0071] The embodiment of the present invention first obtains multiple target monitoring data at different acquisition times at different positions of the target transformer, then uses each target monitoring data as a characteristic dimension of the target transformer, and constructs a multidimensional tensor according to the time dimension, space dimension and characteristic dimension of the multiple target monitoring data; then the multidimensional tensor is divided according to the time dimension, space dimension and characteristic dimension respectively, to obtain the space dimension and characteristic dimension slice data, the time dimension and characteristic dimension slice data, and the time dimension and space dimension slice data of the target transformer; then the correlation between the space dimension and characteristic dimension slice data, the time dimension and characteristic dimension slice data, and the time dimension and space dimension slice data are extracted respectively, and the fault prediction result of the target transformer is obtained according to the extraction result. Therefore, when obtaining the fault prediction result of the target transformer, the multiple target monitoring data of the target transformer can be integrated, and the correlation between the multiple target monitoring information of the target transformer in each dimension can be fully extracted, thereby improving the accuracy of transformer fault prediction.

[0072] It should be understood that the order of execution of the steps in the above embodiment does not necessarily mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present invention.

[0073] The following is an embodiment of the device of the present invention. For details not described in detail therein, reference may be made to the corresponding method embodiment described above.

[0074] Figure 4 The structure diagram of the transformer fault prediction device provided by the embodiment of the present invention is shown. For the convenience of explanation, only the part related to the embodiment of the present invention is shown, which is described in detail as follows:

[0075] like Figure 4 As shown, the transformer fault prediction device includes: an acquisition module 41 , a first processing module 42 , a second processing module 43 and a prediction module 44 .

[0076] The acquisition module 41 is used to acquire a variety of target monitoring data at different locations of the target transformer at different acquisition times.

[0077] The first processing module 42 is used to use each target monitoring data as a feature dimension of the target transformer, and construct a multi-dimensional tensor of multiple target monitoring data according to the time dimension, space dimension and feature dimension.

[0078] The second processing module 43 is used to split the multidimensional tensor according to the time dimension, space dimension and feature dimension respectively, to obtain the space dimension and feature dimension slice data, the time dimension and feature dimension slice data and the time dimension and space dimension slice data of the target transformer.

[0079] The prediction module 44 is used to extract the correlation between the spatial dimension and feature dimension slice data, the time dimension and feature dimension slice data, and the time dimension and spatial dimension slice data, and obtain the fault prediction result of the target transformer according to the extraction result.

[0080] The embodiment of the present invention first obtains multiple target monitoring data at different acquisition times at different positions of the target transformer, then uses each target monitoring data as a characteristic dimension of the target transformer, and constructs a multidimensional tensor according to the time dimension, space dimension and characteristic dimension of the multiple target monitoring data; then the multidimensional tensor is divided according to the time dimension, space dimension and characteristic dimension respectively, to obtain the space dimension and characteristic dimension slice data, the time dimension and characteristic dimension slice data, and the time dimension and space dimension slice data of the target transformer; then the correlation between the space dimension and characteristic dimension slice data, the time dimension and characteristic dimension slice data, and the time dimension and space dimension slice data are extracted respectively, and the fault prediction result of the target transformer is obtained according to the extraction result. Therefore, when obtaining the fault prediction result of the target transformer, the multiple target monitoring data of the target transformer can be integrated, and the correlation between the multiple target monitoring information of the target transformer in each dimension can be fully extracted, thereby improving the accuracy of transformer fault prediction.

[0081] In a possible implementation, the prediction module 44 is specifically configured to:

[0082] Extract the correlation between the spatial dimension and the feature dimension slice data, and record it as the first correlation. ;

[0083] The correlation between the slice data of the time dimension and the feature dimension is extracted and recorded as the second correlation.

[0084] Extract the correlation between the slice data of the time dimension and the space dimension, which is recorded as the third correlation;

[0085] The first correlation, the second correlation and the third correlation are input into the fully connected network to obtain the fault prediction result of the target transformer.

[0086] In a possible implementation, the prediction module 44 is specifically used to extract the correlation between the spatial dimension and the feature dimension slice data based on the graph convolutional network, which is recorded as the first correlation.

[0087] In a possible implementation, the prediction module 44 is specifically configured to extract the correlation between the slice data of the time dimension and the feature dimension based on the gated recurrent unit, which is recorded as the second correlation.

[0088] In a possible implementation, the prediction module 44 is specifically used to extract the correlation between the slice data of the time dimension and the space dimension based on the spatiotemporal graph convolutional network, which is recorded as the third correlation.

[0089] In a possible implementation, the acquisition module 41 can also be used to perform data cleaning and data standardization on each target monitoring data to obtain preprocessed target monitoring data; the first processing module 42 is specifically used to: construct a multidimensional tensor of multiple preprocessed target monitoring data according to the time dimension, space dimension and feature dimension.

[0090] Figure 5 is a schematic diagram of a terminal provided by an embodiment of the present invention. Figure 5 As shown, the terminal 5 of this embodiment includes: a processor 50, a memory 51, and a computer program 52 stored in the memory 51 and executable on the processor 50. When the processor 50 executes the computer program 52, the steps in the above-mentioned transformer fault prediction method embodiments are implemented, for example Figure 1 Alternatively, when the processor 50 executes the computer program 52, the functions of each module / unit in the above-mentioned device embodiments are implemented, for example Figure 4 The functions of the modules 41 to 44 are shown.

[0091] Exemplarily, the computer program 52 may be divided into one or more modules / units, one or more modules / units are stored in the memory 51 and executed by the processor 50 to implement the present invention. The one or more modules / units may be a series of computer program instruction segments capable of implementing specific functions, and the instruction segments are used to describe the execution process of the computer program 52 in the terminal 5. For example, the computer program 52 may be divided into Figure 4 Modules 41 to 44 are shown.

[0092] The terminal 5 may be a computing device such as a desktop computer, a notebook, a PDA, or a cloud server. The terminal 5 may include, but is not limited to, a processor 50 and a memory 51. Those skilled in the art will appreciate that Figure 5 It is only an example of terminal 5 and does not constitute a limitation on terminal 5. It may include more or fewer components than shown in the figure, or a combination of certain components, or different components. For example, the terminal may also include input and output devices, network access devices, buses, etc.

[0093] The processor 50 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc.

[0094] The memory 51 may be an internal storage unit of the terminal 5, such as a hard disk or memory of the terminal 5. The memory 51 may also be an external storage device of the terminal 5, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc. equipped on the terminal 5. Further, the memory 51 may also include both an internal storage unit of the terminal 5 and an external storage device. The memory 51 is used to store computer programs and other programs and data required by the terminal. The memory 51 may also be used to temporarily store data that has been output or is to be output.

[0095] The technicians in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0096] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0097] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0098] In the embodiments provided by the present invention, it should be understood that the disclosed devices / terminals and methods can be implemented in other ways. For example, the device / terminal embodiments described above are only schematic, for example, the division of modules or units is only a logical function division, and there may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0099] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0100] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0101] If the integrated module / unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present invention implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by the processor, the steps of the above-mentioned transformer fault prediction method embodiments can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device capable of carrying computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal and software distribution medium.

[0102] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention, and should be included in the protection scope of the present invention.

Claims

1. A transformer fault prediction method, characterized in that: include: Acquire multiple target monitoring data at different locations of the target transformer at different collection times; Taking each target monitoring data as a feature dimension of the target transformer, constructing a multidimensional tensor of the multiple target monitoring data according to the time dimension, space dimension and feature dimension; The multidimensional tensor is divided according to the time dimension, the space dimension and the feature dimension respectively to obtain the space dimension and feature dimension slice data, the time dimension and feature dimension slice data and the time dimension and space dimension slice data of the target transformer; The spatial dimension and feature dimension slice data, the time dimension and feature dimension slice data, and the correlation between the time dimension and spatial dimension slice data are extracted respectively, and the fault prediction result of the target transformer is obtained according to the extraction result.

2. The transformer fault prediction method according to claim 1, characterized in that: Respectively extracting the spatial dimension and feature dimension slice data, the time dimension and feature dimension slice data, and the correlation between the time dimension and spatial dimension slice data, and obtaining the fault prediction result of the target transformer according to the extraction result, including: Extracting the correlation between the spatial dimension and the feature dimension slice data, recorded as the first correlation; Extracting the correlation between the slice data of the time dimension and the feature dimension, and recording it as the second correlation; Extracting the correlation between the slice data of the time dimension and the space dimension, recorded as the third correlation; The first correlation, the second correlation and the third correlation are input into a fully connected network to obtain a fault prediction result of the target transformer.

3. The transformer fault prediction method according to claim 2, characterized in that: Extracting the correlation between the spatial dimension and the feature dimension slice data, recorded as the first correlation, includes: The correlation between the spatial dimension and feature dimension slice data is extracted based on the graph convolutional network and recorded as the first correlation.

4. The transformer fault prediction method according to claim 2, characterized in that: Extracting the correlation between the time dimension and the feature dimension slice data, recorded as the second correlation, includes: The correlation between the time dimension and feature dimension slice data is extracted based on the gated recurrent unit and recorded as the second correlation.

5. The transformer fault prediction method according to claim 2, characterized in that: Extracting the correlation between the slice data of the time dimension and the space dimension, recorded as the third correlation, includes: The correlation between the slice data of the time dimension and the space dimension is extracted based on the spatiotemporal graph convolutional network and recorded as the third correlation.

6. The transformer fault prediction method according to claim 1, characterized in that: After obtaining a variety of target monitoring data at different locations of the target transformer at different acquisition times, it also includes: Performing data cleaning and data standardization on each of the target monitoring data to obtain pre-processed target monitoring data; The target monitoring data are constructed into a multidimensional tensor according to the time dimension, space dimension and feature dimension, including: A variety of pre-processed target monitoring data are constructed into multi-dimensional tensors according to the time dimension, space dimension and feature dimension.

7. A transformer fault prediction device, characterized in that: include: An acquisition module, used to acquire a variety of target monitoring data at different locations of a target transformer at different acquisition times; A first processing module is used to use each of the target monitoring data as a feature dimension of the target transformer, and construct a multidimensional tensor of the multiple target monitoring data according to the time dimension, space dimension and feature dimension; The second processing module is used to slice the multidimensional tensor according to the time dimension, the space dimension and the feature dimension respectively, to obtain the space dimension and feature dimension slice data, the time dimension and feature dimension slice data and the time dimension and space dimension slice data of the target transformer; A prediction module is used to respectively extract the spatial dimension and feature dimension slice data, the time dimension and feature dimension slice data, and the correlation between the time dimension and spatial dimension slice data, and obtain a fault prediction result of the target transformer based on the extraction result.

8. The transformer fault prediction device according to claim 7, characterized in that: The prediction module is specifically used for: Extracting the correlation between the spatial dimension and the feature dimension slice data, recorded as the first correlation; Extracting the correlation between the slice data of the time dimension and the feature dimension, and recording it as the second correlation; Extracting the correlation between the slice data of the time dimension and the space dimension, recorded as the third correlation; The first correlation, the second correlation and the third correlation are input into a fully connected network to obtain a fault prediction result of the target transformer.

9. A terminal, characterized in that: The method comprises a memory and a processor, wherein the memory is used to store a computer program, and the processor is used to call and run the computer program stored in the memory to execute the method according to any one of claims 1 to 6.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images